Artificial Intelligence (AI) in natural drug discovery
Start with the big picture
Machine learning, deep learning, and natural-language processing support different stages of NP discovery. Models can predict structure–activity relationships and rank candidate compounds, while spectral analysis can help distinguish known molecules from potential new leads. Genomic pipelines identify biosynthetic gene clusters, and generative approaches propose NP analogues for evaluation. Virtual screening, network pharmacology, and integration with omics data extend analysis from individual compounds to targets, pathways, and phenotypic effects. The workflow also includes early in silico assessment of ADMET and toxicity, with experimental results feeding back into active-learning cycles. The deeper lesson examines these applications alongside supporting resources such as AI-enhanced NP databases, graph neural networks, explainable AI, and the challenges and future directions of the field.
What you'll learn
- Describe how AI supports mining of natural-product datasets and prioritization of candidate bioactives.
- Distinguish key uses of machine learning, deep learning, and natural-language processing in NP discovery.
- Explain how dereplication and genomic AI pipelines contribute to compound discovery.
- Outline the roles of virtual screening, network pharmacology, and omics integration.
- Identify how ADMET prediction and active learning fit into the discovery workflow.
Continue your study
Work through the complete notes and reinforce the topic with the study tools available in the full lesson.
